CN-122023415-A - Bridge girder crack detection method and system based on unmanned aerial vehicle

CN122023415ACN 122023415 ACN122023415 ACN 122023415ACN-122023415-A

Abstract

A bridge girder crack detection method and system based on unmanned aerial vehicle belongs to the technical field of bridge detection. The method comprises the steps of calculating a spatial view factor of an unmanned aerial vehicle, mapping a crack identification result of the bridge into a continuous two-dimensional spatial distribution field by adopting a probability diffusion mode to obtain crack spatial distribution probability and crack tensor, converting a determinant of a trace item and the crack tensor based on the crack tensor into a thermal potential function, then constructing a crack driving force distribution field reflecting a crack pushing trend, constructing a crack evolution model by adopting a reaction-diffusion equation mode, solving final evolution stable crack density, quantifying the complexity degree of the crack probability at each spatial position by adopting an information entropy density, then establishing regional risk indexes by utilizing the information entropy density pair, the final evolution stable crack density and the crack driving force, and carrying out grading early warning and response. The method improves the prediction capability of the bridge girder crack detection.

Inventors

LIU XING
CHENG GONG
LI JUNYUAN
MENG ANXIN
Ren Bangke
FENG JUNHUA

Assignees

深城交科技集团股份有限公司

Dates

Publication Date: 20260512
Application Date: 20260414

Claims (8)

1. The bridge girder crack detection method based on the unmanned aerial vehicle is characterized by comprising the following steps of: s1, when detecting surface cracks of a bridge structure, taking shooting angles and shooting distance factors of a plurality of unmanned aerial vehicles into consideration, carrying out multi-angle space adaptability analysis on the unmanned aerial vehicles, and calculating space visual angle factors of the unmanned aerial vehicles; S2, mapping a crack identification result of the bridge into a continuous two-dimensional space distribution field by adopting a probability diffusion mode based on the space visual angle factor of the unmanned aerial vehicle obtained in the step S1 to obtain the crack space distribution probability and the crack tensor; S3, converting the crack tensor obtained in the step S2 into a thermal potential function based on a trace item of the crack tensor and a determinant of the crack tensor, and then constructing a crack driving force distribution field reflecting a crack pushing trend; s4, constructing a crack evolution model by using a reaction-diffusion equation form based on the crack driving force distribution field obtained in the step S3, and solving the final evolution stable crack density; S5, quantifying the complexity degree of the crack probability of each spatial position by adopting information entropy density, and then establishing an area risk index by utilizing the information entropy density pair, the final evolution stable crack density and the crack driving force; And S6, carrying out grading early warning and response on the regional risk index obtained in the step S5, and completing bridge girder crack detection based on the unmanned aerial vehicle.
2. The bridge girder crack detection method based on the unmanned aerial vehicle according to claim 1, wherein the expression for calculating the spatial viewing angle factor of the unmanned aerial vehicle in step S1 is: ; Wherein, the Slit pixel point coordinates for nth image N is the image number, K is the unmanned aerial vehicle number, and K is the total number of unmanned aerial vehicles; for shooting weight factors, determining by an expert in combination with angle errors and image resolution factors, and simultaneously adjusting dimensions; The method is characterized in that the method comprises the steps of acquiring a space position of a kth unmanned aerial vehicle by a GNSS system, wherein x is a coordinate of a crack pixel point along the length direction of a bridge girder, y is a coordinate of the crack pixel point along the width direction of the bridge girder, h is the height of the bridge girder surface where the crack pixel point is located, and the height is acquired by a design drawing.
3. The bridge girder crack detection method based on the unmanned aerial vehicle according to claim 2, wherein the specific implementation method of the step S2 comprises the following steps: S2.1, extracting crack pixel points and confidence coefficient of the unmanned aerial vehicle image by means of an image recognition algorithm, and performing diffusion modeling on the crack pixel points according to a probability method to obtain crack space distribution probability, wherein the expression is as follows: ; Wherein, the Slit pixel point coordinates for nth image I is the number of the pixel points of the crack; The number of split pixel points in the nth image; Calling an image recognition algorithm for the confidence coefficient of the crack pixel point to obtain; The pixel point i of the crack is the coordinate along the length direction of the bridge girder and the coordinate along the width direction of the bridge girder; is a space fuzzy scale and is determined by expert experience; Slit pixel point coordinates for nth image A corresponding spatial viewing angle factor; s2.2, constructing a crack tensor based on the crack space distribution probability, wherein the expression is as follows: ; Wherein, the Slit pixel point coordinates for nth image A corresponding fracture tensor; the probability variation of the crack space distribution in the x direction of the nth image; the probability variation of the crack space distribution in the y direction of the nth image.
4. The bridge girder crack detection method based on the unmanned aerial vehicle according to claim 3, wherein the specific implementation method of the step S3 comprises the following steps: s3.1, constructing a thermal potential function, wherein the expression is: ; Wherein, the Slit pixel point coordinates for nth image A corresponding thermal potential function; a mathematical manipulation tensor trace; is a mathematical manipulation tensor determinant; Is a weighting coefficient and is determined empirically by an expert; s3.2, taking a thermal potential function as a potential field for driving crack development, and obtaining a driving force distribution field of the crack by gradient of the thermal potential function, wherein the expression is as follows: ; Wherein, the Slit pixel point coordinates for nth image Corresponding crack driving force; gradient operators are operated on mathematically.
5. The unmanned aerial vehicle-based bridge girder crack detection method according to claim 4, wherein the specific implementation method of the step S4 comprises the following steps: s4.1, constructing a crack evolution model, wherein the expression is as follows: ; Wherein, the The crack density of a space region where an nth image is positioned in an evolution process; Is the evolution time; For simulating the diffusion coefficient, determining according to the bridge girder material by expert experience; for dynamic response enhancement, determined by simulation or expert experience; a mold length which is a crack driving force of the nth image; s4.2, based on the fact that the numerical solution reaches stable convergence in the evolution process and is updated, the final evolution stable crack density is solved, and the solving process is as follows: ; ; ; Wherein, the Slit pixel point coordinates for nth image Corresponding initial crack density; Is a time interval and is determined empirically by an expert; Is a preset tolerance; Slit pixel point coordinates for nth image The corresponding crack density in the evolution of step b; stabilizing crack density for final evolution of the nth image; the final crack density for the nth image.
6. The unmanned aerial vehicle-based bridge girder crack detection method according to claim 5, wherein the specific implementation method of the step S5 comprises the following steps: S5.1, establishing information entropy density, wherein the expression is: ; ; Wherein, the Slit pixel point coordinates for nth image A corresponding information entropy density; For penalty term components, determined empirically by an expert; Slit pixel point coordinates for nth image Corresponding multi-view coverage times are determined by data collected by the unmanned aerial vehicle; representing an effective visual coverage area photographed by a kth unmanned aerial vehicle; S5.2. Through 、、 The three indexes are fused, and a risk index is constructed in a weighted integral form of the region, and the expression is as follows: ; Wherein, the To designate a spatial region; Designating a region risk index in a spatial region for an nth image; 、、 the fracture density factor, the driving direction factor and the information complexity factor are respectively determined by expert experience.
7. The unmanned aerial vehicle-based bridge girder crack detection method according to claim 6, wherein the step S6 of grading early warning and response method comprises the following steps: When (when) When the evaluation area is stable, no additional engineering treatment is needed, and the classification is L1; When (when) When the situation that the crack is potentially expanded exists in the evaluation area, the situation needs to be reviewed and daily maintenance work is carried out on the evaluation area, and the evaluation area is classified into a class L2; When (when) When the crack propagation trend of the evaluation area is obvious, maintenance and reinforcement are required, and the evaluation area is classified into a grade L3; When (when) When the crack expansion speed of the evaluation area is relatively high, the closed early warning is carried out, and the evaluation area is classified into a class L4; 、、 The boundary values of the stable region and the potential risk, the boundary values of the potential region and the developing risk and the high risk critical value are respectively obtained.
8. A system based on an unmanned aerial vehicle bridge girder crack detection method, comprising a processor, a memory and a computer program stored in the memory and operable on the processor, the computer program when run implementing the steps of an unmanned aerial vehicle bridge girder crack detection method as claimed in any one of claims 1 to 7.

Description

Bridge girder crack detection method and system based on unmanned aerial vehicle Technical Field The invention belongs to the technical field of bridge detection, and particularly relates to a bridge girder crack detection method and system based on an unmanned aerial vehicle. Background In a bridge structural system, a main beam is a key part for bearing, the health condition of the main beam has very key influence on the safety and the service life of the whole structure, the main beam can be subjected to various loads and environmental factors in the long-term working process, cracks are easy to appear at the parts of a web plate, a lower edge and a span, the cracks are often characterized in a larger direction, the structure is complex, if the cracks cannot be timely perceived and treated, the structure is possibly unstable, or the using function is reduced, so that the main beam needs to be continuously monitored, and accurate judgment is also carried out. In terms of monitoring means, the traditional mode of manual detection or ground equipment observation is difficult to fully consider key parts such as the upper surface and the lower surface of the main beam because of insufficient visual angle and more shielding. The unmanned aerial vehicle is gradually guided into for auxiliary detection in the years, and the defects of ground perception are supplemented by utilizing the advantages of flexible flight path and multi-angle observation of the unmanned aerial vehicle, so that the acquired information is more complete. But the current evaluation method is relatively lacking in deep analysis and predictive ability of overall crack evolution trend. Disclosure of Invention The invention aims to solve the problem of realizing accurate assessment of the risk level of the crack of the bridge girder, and provides a bridge girder crack detection method and system based on an unmanned aerial vehicle. In order to achieve the above purpose, the present invention is realized by the following technical scheme: A bridge girder crack detection method based on an unmanned aerial vehicle comprises the following steps: s1, when detecting surface cracks of a bridge structure, taking shooting angles and shooting distance factors of a plurality of unmanned aerial vehicles into consideration, carrying out multi-angle space adaptability analysis on the unmanned aerial vehicles, and calculating space visual angle factors of the unmanned aerial vehicles; S2, mapping a crack identification result of the bridge into a continuous two-dimensional space distribution field by adopting a probability diffusion mode based on the space visual angle factor of the unmanned aerial vehicle obtained in the step S1 to obtain the crack space distribution probability and the crack tensor; S3, converting the crack tensor obtained in the step S2 into a thermal potential function based on a trace item of the crack tensor and a determinant of the crack tensor, and then constructing a crack driving force distribution field reflecting a crack pushing trend; s4, constructing a crack evolution model by using a reaction-diffusion equation form based on the crack driving force distribution field obtained in the step S3, and solving the final evolution stable crack density; S5, quantifying the complexity degree of the crack probability of each spatial position by adopting information entropy density, and then establishing an area risk index by utilizing the information entropy density pair, the final evolution stable crack density and the crack driving force; And S6, carrying out grading early warning and response on the regional risk index obtained in the step S5, and completing bridge girder crack detection based on the unmanned aerial vehicle. Further, in step S1, the expression for calculating the spatial viewing angle factor of the unmanned aerial vehicle is: Wherein, the Slit pixel point coordinates for nth imageN is the image number, K is the unmanned aerial vehicle number, and K is the total number of unmanned aerial vehicles; for shooting weight factors, determining by an expert in combination with angle errors and image resolution factors, and simultaneously adjusting dimensions; The method is characterized in that the method comprises the steps of acquiring a space position of a kth unmanned aerial vehicle by a GNSS system, wherein x is a coordinate of a crack pixel point along the length direction of a bridge girder, y is a coordinate of the crack pixel point along the width direction of the bridge girder, h is the height of the bridge girder surface where the crack pixel point is located, and the height is acquired by a design drawing. Further, the specific implementation method of the step S2 includes the following steps: S2.1, extracting crack pixel points and confidence coefficient of the unmanned aerial vehicle image by means of an image recognition algorithm, and performing diffusion modeling on the crack pixel points according to a pr